1 | package manifest;
|
---|
2 |
|
---|
3 |
|
---|
4 | use XMLParser;
|
---|
5 | use strict;
|
---|
6 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
7 |
|
---|
8 |
|
---|
9 | our $self;
|
---|
10 |
|
---|
11 | sub new {
|
---|
12 | my ($class) = shift (@_);
|
---|
13 |
|
---|
14 | $self = {} ;
|
---|
15 |
|
---|
16 | $self->{'index'} = {};
|
---|
17 | $self->{'reindex'} = {};
|
---|
18 | $self->{'delete'} = {};
|
---|
19 |
|
---|
20 | return bless $self, $class;
|
---|
21 | }
|
---|
22 |
|
---|
23 | sub parse
|
---|
24 | {
|
---|
25 | my ($self) = shift (@_);
|
---|
26 | my ($filename) = @_;
|
---|
27 |
|
---|
28 | my $parser = new XML::Parser('Style' => 'Stream',
|
---|
29 | 'Handlers' => {'Char' => \&Char,
|
---|
30 | 'XMLDecl' => \&XMLDecl,
|
---|
31 | 'Entity' => \&Entity,
|
---|
32 | 'Doctype' => \&Doctype,
|
---|
33 | 'Default' => \&Default
|
---|
34 | });
|
---|
35 |
|
---|
36 | $parser->parsefile($filename);
|
---|
37 | }
|
---|
38 |
|
---|
39 | sub StartDocument {$self->xml_start_document(@_);}
|
---|
40 | sub XMLDecl {$self->xml_xmldecl(@_);}
|
---|
41 | sub Entity {$self->xml_entity(@_);}
|
---|
42 | sub Doctype {$self->xml_doctype(@_);}
|
---|
43 | sub StartTag {$self->xml_start_tag(@_);}
|
---|
44 | sub EndTag {$self->xml_end_tag(@_);}
|
---|
45 | sub Text {$self->xml_text(@_);}
|
---|
46 | sub PI {$self->xml_pi(@_);}
|
---|
47 | sub EndDocument {$self->xml_end_document(@_);}
|
---|
48 | sub Default {$self->xml_default(@_);}
|
---|
49 |
|
---|
50 | # This Char function overrides the one in XML::Parser::Stream to overcome a
|
---|
51 | # problem where $expat->{Text} is treated as the return value, slowing
|
---|
52 | # things down significantly in some cases.
|
---|
53 | sub Char {
|
---|
54 | use bytes; # Necessary to prevent encoding issues with XML::Parser 2.31+
|
---|
55 | $_[0]->{'Text'} .= $_[1];
|
---|
56 | return undef;
|
---|
57 | }
|
---|
58 |
|
---|
59 | # Called at the beginning of the XML document.
|
---|
60 | sub xml_start_document {
|
---|
61 | my $self = shift(@_);
|
---|
62 | my ($expat) = @_;
|
---|
63 |
|
---|
64 | }
|
---|
65 |
|
---|
66 | # Called for XML declarations
|
---|
67 | sub xml_xmldecl {
|
---|
68 | my $self = shift(@_);
|
---|
69 | my ($expat, $version, $encoding, $standalone) = @_;
|
---|
70 | }
|
---|
71 |
|
---|
72 | # Called for XML entities
|
---|
73 | sub xml_entity {
|
---|
74 | my $self = shift(@_);
|
---|
75 | my ($expat, $name, $val, $sysid, $pubid, $ndata) = @_;
|
---|
76 | }
|
---|
77 |
|
---|
78 | # Called for DOCTYPE declarations - use die to bail out if this doctype
|
---|
79 | # is not meant for this plugin
|
---|
80 | sub xml_doctype {
|
---|
81 | my $self = shift(@_);
|
---|
82 | my ($expat, $name, $sysid, $pubid, $internal) = @_;
|
---|
83 | die "Manifest Cannot process XML document with DOCTYPE of $name";
|
---|
84 | }
|
---|
85 |
|
---|
86 | # Called for every start tag. The $_ variable will contain a copy of the
|
---|
87 | # tag and the %_ variable will contain the element's attributes.
|
---|
88 | sub xml_start_tag
|
---|
89 | {
|
---|
90 | my $self = shift(@_);
|
---|
91 | my ($expat, $element) = @_;
|
---|
92 |
|
---|
93 | if ($element eq "Filename")
|
---|
94 | {
|
---|
95 | $self->{'filename'} = "";
|
---|
96 | }
|
---|
97 | else
|
---|
98 | {
|
---|
99 | if (defined($self->{'file-type'}))
|
---|
100 | {
|
---|
101 | print STDERR "Warning: Malformed XML manifest ($element nested inside " . $self->{'file-type'} . ")\n";
|
---|
102 | }
|
---|
103 |
|
---|
104 | $self->{'file-type'} = $element;
|
---|
105 | }
|
---|
106 | }
|
---|
107 |
|
---|
108 | # Called for every end tag. The $_ variable will contain a copy of the tag.
|
---|
109 | sub xml_end_tag
|
---|
110 | {
|
---|
111 | my $self = shift(@_);
|
---|
112 | my ($expat, $element) = @_;
|
---|
113 |
|
---|
114 | if ($element eq "Filename")
|
---|
115 | {
|
---|
116 | $self->{lc($self->{'file-type'})}->{$self->{'filename'}} = 1;
|
---|
117 | $self->{'filename'} = undef;
|
---|
118 | }
|
---|
119 | else
|
---|
120 | {
|
---|
121 | $self->{'file-type'} = undef;
|
---|
122 | }
|
---|
123 | }
|
---|
124 |
|
---|
125 | # Called just before start or end tags with accumulated non-markup text in
|
---|
126 | # the $_ variable.
|
---|
127 | sub xml_text {
|
---|
128 | my $self = shift(@_);
|
---|
129 | my ($expat) = @_;
|
---|
130 |
|
---|
131 | if (defined $self->{'filename'}) {
|
---|
132 | my $text = $_;
|
---|
133 | chomp($text);
|
---|
134 |
|
---|
135 | $text =~ s/^\s+//;
|
---|
136 | $text =~ s/\s+$//;
|
---|
137 |
|
---|
138 | $self->{'filename'} .= $text if ($text !~ m/^\s*$/);
|
---|
139 | }
|
---|
140 | }
|
---|
141 |
|
---|
142 | # Called for processing instructions. The $_ variable will contain a copy
|
---|
143 | # of the pi.
|
---|
144 | sub xml_pi {
|
---|
145 | my $self = shift(@_);
|
---|
146 | my ($expat, $target, $data) = @_;
|
---|
147 | }
|
---|
148 |
|
---|
149 | # Called at the end of the XML document.
|
---|
150 | sub xml_end_document {
|
---|
151 | my $self = shift(@_);
|
---|
152 | my ($expat) = @_;
|
---|
153 |
|
---|
154 | }
|
---|
155 |
|
---|
156 | # Called for any characters not handled by the above functions.
|
---|
157 | sub xml_default {
|
---|
158 | my $self = shift(@_);
|
---|
159 | my ($expat, $text) = @_;
|
---|
160 | }
|
---|
161 |
|
---|
162 |
|
---|
163 | 1;
|
---|